-
-
Notifications
You must be signed in to change notification settings - Fork 314
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
RISC-V test plan #4976
Comments
The tests are broken and should be disabled until fixed. Relates to adoptium#4976
The tests are broken and should be disabled until fixed. Relates to adoptium#4976
The tests are broken and should be disabled until fixed. Relates to #4976
From a look over the results of https://ci.adoptium.net/job/Test_openjdk17_hs_extended.openjdk_riscv64_linux/10, I can identify at least two fixes that we need to do in OpenJDK, none are specific to RISC-V:
I still have to identify the cause of the crash for |
Some tests are timing out even though they are not stuck. Increase the timeoutFactor to give them a chance to finish Relates to adoptium#4976
From early debugging, it's most likely related to the message |
|
* Increase timeout on riscv64 Some tests are timing out even though they are not stuck. Increase the timeoutFactor to give them a chance to finish Relates to #4976 * Update openjdk/openjdk.mk Co-authored-by: Martijn Verburg <[email protected]> * Update openjdk/openjdk.mk Co-authored-by: Martijn Verburg <[email protected]> --------- Co-authored-by: Martijn Verburg <[email protected]>
From the latest run on 21.0.2+13:
References:
|
We excluded that perf target for AIX here, it is not unreasonable to exclude for risc-v, though I have not seen 'how it fails' so will be worth understanding before excluding |
Noting that Grinder 9060 seemed to have the philosophers test disabled, and therefore extended.perf passed overall so I guess it was excluded in |
TestJlmRemoteNotifierProxyAuth_0 failed in sanity.system at https://ci.adoptium.net/job/Test_openjdk21_hs_sanity.system_riscv64_linux/56/console but seemed to pass in the automatic re-run at https://ci.adoptium.net/job/Test_openjdk21_hs_sanity.system_riscv64_linux_rerun/9/consoleFull (Both were on the scaleway-1 machine. Grinding with 10 iterations at https://ci.adoptium.net/job/Grinder/9083/ (EDIT: Failed 1/10) testList_2 had a few targets failed in the run above but a subsequent run passed at https://ci.adoptium.net/job/Test_openjdk21_hs_extended.openjdk_riscv64_linux_testList_2/5/console So I believe we're only left with the final bullet point from the earlier comment: Failures in extended.openjdk: TestJcmdDefaults (jdk_tools_1) and VectorMaxConversionTests_Zsinglegen (jdk_vector_[01]) |
VectorMaxConversionTests_Zsinglegen passed in https://ci.adoptium.net/job/Grinder/9111/ on |
This looks good. We should gather the .tap files from test jobs and Grinders and attach to this issue to close. |
TAP Files:
|
not sure why those Grinders would not have tap file artifact, from console of Grinder/9060
and also https://ci.adoptium.net/job/Grinder/9060/tapResults/ indicates it was available to present with the TAP plugin. Also the benchmark_output.tar.gz and testenv.properties file should have been archived but are not present. - should not matter for testenv.properties and .tap files, which should always be archived. |
I've re-run those Grinders (9135, 9136), and collected the additional two tap files: Ref the comment from earlier in this issue: |
Given we have seen all AQAvit and TCK tests passing for this platform, it is 👍 to publish. |
Relates to adoptium/temurin-build#3591
Action Items
gnu.testlet.java.lang.Math
andgnu.testlet.java.lang.StrictMath
in Mauve directly, the tests are broken for platform that have canonical NaNsTest Failures
MiniMix_aot_5m
MiniMix_5m
MauveSingleThrdLoad_HS_5m
MauveSingleInvocLoad_HS_5m
MauveMultiThrdLoad_5m
ConcurrentLoadTest_5m
renaissance-philosophers
The text was updated successfully, but these errors were encountered: